VIROLOGY

186,452-462

Molecular

(1992)

Cloning and Expression

of the Bacteriophage

JOSEPH MICHALEWICZ’ Department

of Biological

Sciences,

AND

T7 0.7(Protein

Kinase) Gene

ALLEN W. NICHOLSON2

Wayne State University,

Received June 3, 199 1; accepted

Detroit, Michigan

48202

October 24, 199 1

The bacteriophage T7 0.7 gene encodes a protein which supports viral reproduction under specific suboptimal growth conditions. The 0.7 protein (gp0.7) shuts off host RNA polymerase-catalyzed transcription and also expresses a serine/threonine-specific, CAMP-independent protein kinase (PK) activity. To determine the role of the gp0.7 PK in viral reproduction, the 0.7 gene of the T7(JS78) mutant phage-whose gp0.7 expresses only the PK activity-was cloned in the plasmid expression vector PET-1 1 a. Cells containing the recombinant plasmid were viable, and upon IPTG induction produced a 30-kDa polypeptide, similar in size to the gp0.7-related polypeptide seen in T7(JS78)-infected cells. Extracts of cells containing this polypeptide can phosphorylate the exogenous substrate lysozyme. Expression of plasmid-encoded gpO.7(JS78) in vivo results in phosphorylation of the same proteins which are phosphorylated in T7(JS78)-infected cells; moreover, the plasmid-encoded gpO.7(JS78) is itself phosphorylated. The JS78 mutation changes Gln243 in gp0.7 to an amber codon, which explains the production of the truncated, 30-kDa gp0.7-related polypeptide, and implicates the 1 1-kDa C-terminal domain in host transcription shut-off. The T7(A23) 0.7 point mutant fails to express PK activity in infected cells. However, the truncated gp0.7(A23)-related polypeptide, expressed from a plasmid, exhibits PK activity in vivo and in vitro, but with an altered specificity. Thus, the A23 mutation, which changes 0 1992 Academic Press. hc. Asp1 00 to Asn, may identify a substrate recognition determinant.

specific and CAMP-independent. Initial studies of the protein kinase prompted the conclusion that over 40 proteins are substrates for gp0.7-catalyzed phosphorylation in viva (Rahmsdorf et a/., 1974; Pai et al., 1975b). It has been shown recently that a specific set of proteins undergoes prominent phosphorylation in response to gp0.7 protein kinase expression in viva (Robertson and Nicholson, 1990), where it was also shown that uninfected cell phosphoproteins (e.g., see Cortay et a/., 1986) as well as proteins phosphorylated in response to the infection process per se are present in the gel electrophoretic patterns. It is possible that a number of the gp0.7 protein kinase-specific phosphoproteins result from an indirect phosphorylation process (e.g., see ltikawa et a/., 1989). Currently, only two proteins have been firmly identified as substrates for the gp0.7 protein kinase. The processing endoribonuclease RNase III undergoes a stimulation in its cleavage activity in vitro upon phosphorylation of a serine residue(s) (Mayer and Schweiger, 1983). The ,f3’subunit of RNA polymerase is phosphorylated at a threonine residue(s) (Zillig et a/., 1975), and there is evidence for a partial reduction in transcriptional activity in vitro of phosphorylated enzyme (Hesselbach and Nakada, 1977). The reduction of activity in vitro may be related to the gp0.7 protein kinasedependent, enforced termination of host transcription in vivo at otherwise weak terminator signals (Ponta et al., 1974; Pfennig-Yeh et al., 1978). As there is now evidence for the involvement of the p’ subunit in tran-

INTRODUCTION The bacteriophage T7 expresses from its genetic early region approximately 10 polypeptides, at least several of which are involved in converting the infected host cell environment to one optimal for efficient viral reproduction (Kruger and Schroeder, 1981; Dunn and Studier, 1983; Studier and Dunn, 1983; Hausmann, 1988). In one set of early events, the pattern of gene expression in the T7-infected cell quickly shifts from cell- to viral-directed, as the gene 1-encoded RNA polymerase is synthesized, and transcription of the phage Class II and III genes is initiated. This shift is concurrently reinforced by the shut off of host RNA polymerase-catalyzed transcription, which is exerted by the product of the 0.7 gene (gp0.7) (Brunovskis and Summers, 1971, 1972; Rothman-Denes eta/., 1973). Functions of the other T7 early gene products include an inhibitor of Type I restriction endonucleases (gp0.3) (Studier, 1975) an inhibitor of the cell-encoded dGTP triphosphohydrolase (gpl.2) (Huber eta/., 1988) and a DNA ligase activity (gpl.3) (Masamune et a/., 1971) (see Fig. 1). The biological activities and functional roles of the other phage early proteins remain obscure. The T7 gp0.7 also expresses a protein kinase activity (Rahmsdorf et al., 1974) which is serine/threonine’ Present address: Natural Sciences and Mathematics ment, Holy Family College, Philadelphia, PA 191 14. ’ To whom requests for reprints should be addressed. 0042-6822/92

$3.00

Copynght 0 1992 by Academic Press. Inc All rights of reproduction !n any form reserved.

Depart-

452

BACTERIOPHAGE ANTI-RESTRICTKX,

PROTEIN KINASE I SHVT OFF

T7 PROTEIN KINASE RNA WLYMERASE

453 dOTPar

INHlSrTOR

DNA LIGASE

FIG. 1. Genetic early region of T7 phage. See Dunn and Studier (1981, 1983) for a physical and genetic description of the T7 genome. The 10 early genes are indicated by the stippled bars; “DR” refers to the direct repeat sequence, and T, denotes the termination signal for the early transcription unit. The host promoters (A,-A,) are indicated, as are the 5 RNase III processing signals (e.g., R0.3). The numbers beneath the linear map refer to position (in kbp) in the nucleotide sequence. Shown beneath the 0.7 gene are the approximate locations of three T7 0.7 gene mutants (A23, JS78, Cl 13) that affect the protein kinase and shut-off activities of gp0.7 (see text).

scription termination (Ito eta/., 1991) perhaps p’phosphorylation stimulates the termination activity of this subunit. However, p’ phosphorylation is not responsible for host shut off, as mutant versions of gp0.7 have been characterized which do not exhibit phosphotransferase activity in v&o, yet can efficiently catalyze shut off (Rothman-Denes et a/., 1973; Simon and Studier, 1973). There is currently limited information on the function(s) of the gp0.7 protein kinase in the T7 reproductive process, but several lines of evidence point to an important role for this activity. First, although the 0.7 gene is dispensible for phage growth under normal laboratory conditions, expression of the gp0.7 protein kinase strongly enhances burst sizes and plating efficiencies when T7 is propagated at elevated temperature, in limiting nutrients, or in the presence of specific phage-inhibitory plasmids (Hirsch-Kaufmann et al., 1975; Gomez and Nualart, 1977). Second, the protein kinase is a conserved function, as it is expressed by all members of the T7 phage group which have been carefully studied (Studier and Mowa, 1976; Studier, 1979; Mertens and Hausmann, 1982; Hodgson et a/., 1985). Third, as mentioned above, a specific set of proteins are phosphorylated in response to gp0.7 protein kinase expression in viva, and of the two gp0.7 protein kinase substrates whose phosphorylated forms have been biochemically characterized (see above), both undergo an alteration (stimulation) in their biological activities upon phosphorylation. A molecular clone of the 0.7 gene would provide new approaches to studying the function and mechanism of the gp0.7 protein kinase. In particular, the controlled expression of gp0.7 from a plasmid would allow an assessment of the effects of gp0.7-catalyzed protein phosphorylation on host gene expression and metabolism, in a manner independent of the viral infection process. A bacterial strain overexpressing gp0.7 would also facilitate purification of homogeneous enzyme for enzymological and structural studies. We describe in this report the molecular cloning of a specific

mutant 0.7 gene, encoding a gp0.7-related polypeptide which expresses the protein kinase, but lacks shut-off activity. The recombinant PET-1 1a plasmid carrying the mutant 0.7 gene expresses, in an IPTG3dependent manner, a 30-kDa polypeptide which contains the N-terminal 68% of gp0.7. Expression of the gp0.7-related polypeptide from the recombinant plasmid in viva results in the phosphorylation of the same proteins as those modified in the phage-infected cell. We also show that the gp0.7-related polypeptide is phosphorylated in viva, and that only minor levels of transphosphorylation activity are observed when the gp0.7-related polypeptide is present in large amounts in its phosphorylatedform. Finally, DNAsequenceanalysis of two 0.7 gene mutants permits a description of how one mutation specifically alters the gp0.7 PK activity, and how the other mutation suppresses gp0.7-catalyzed host shut off. MATERIALS

AND METHODS

Chemicals Chemicals and reagents used were of the highest grade commercially available. The radiolabeled compound [T-~‘P]ATP (6000 Ci/mmol), carrier-free inorganic [32P]phosphate (8500-9120 Ci/mmol), and [35S]Express protein labeling mix (>lOOO Ci/mmol) were obtained from DuPont-NEN (Boston, MA). IPTG was obtained from Sigma (St. Louis, MO). Plasmid PET-1 1a was purchased from Novagen (Madison, WI). T4 DNA ligase and restriction endonucleases Noel and BamHl were from New England Biolabs (Beverly, MA) or Promega (Madison, WI). Sequenase (version 2.0) was obtained from United States Biochemicals (Cleveland, OH), and Vent DNA polymerase was from New

3 Abbreviations used: WT, wild-type; IPTG, Isopropyl-B-o-thiogalactoside; bp, base-pair; m.o.i., multiplicity of infection: p.i., postinfection; kDa, kilodaltons; PCR, polymerase chain reaction; PAGE, polyacrylamide gel electrophoresis; SDS, sodium dodecyl sulfate.

454

MICHALEWICZ

England Biolabs. The enzymes were used according to the instructions supplied by the vendors. DNA oligonucleotides were prepared using an Applied Biosystems Model 3806 DNA synthesizer, and following full deprotection, purified by electrophoresis in 15% polyacrylamide gels containing 7 l\/I urea (Atkinson and Smith, 1984). The deoxyoligonucleotide primers used for PCR-mediated amplification and cloning of the entire T7 0.7(JS78) gene were: Oligo 1, 5’-GGGCATmAACAl-TACCGACATCATG-3’ (2041), containing an Ndel site 4 nt from the 5’ end; and Oligo 2,5’-CCCGGATCCTCAGCCCATTAACATTGC-3’ (3083), containing a f3amHl site 4 nt from the 5’ end [the numbers in parentheses refer to the position in the T7 genome of the 3’ terminal nucleotide (Dunn and Studier, 1983), and the underlined nucleotides in Oligo 1 correspond to the initiator codon for gpO.71. The primers used for amplification and cloning of the approximately 730 bp gp0.7 protein kinase domain of the T7(JS78) and T7(A23) 0.7 genes were Oligo 1 and Oligo 3: 5’-CCCGGATCCCTATCGTGCGACTTCCTC-3’ (2732). Primers used for sequence analysis of the T7(A23) mutation were A23-5’,5’-CAAGATGCTCGGTAATGGTC-3’ (2231-2250) and A23-3’, 5’-TAGCATCCAGCGTGGCGCTG-3’ (2392-2411) (the numbers in parentheses refer to the position in the T7 chromosome of the 5’ and 3’ end nucleotides). The corresponding primers for analysis of the T7(JS78) mutation were JS-5’, 5’-CGACCCGGTATCA‘TTCTCGC-3’(266 l2680) and JS-3’, 5’-CCCATCGCCACGCAGCAAGC-3’ (2867-2886). Bacteria and phage The Escherichia co/i strain W31 10 was obtained from B. Bachmann (E co/i Genetic Stock Center, New Haven, CT), while HMS174 [r; ml recA1 RifR Su-] (Campbell et a/., 1978) and its derivative HMS174(DE3), containing the T7 RNA polymerase gene under control of the lacUV5 promoter in a defective A prophage (Studier et al., 1990), were obtained from Novagen (Madison, WI). T7(WT), and the T7(0.7) mutants T7(A23) and T7(JS78)(Studier, 1973a) were generously provided by F. W. Studier(Brookhaven National Laboratories, Upton, NY). Growth and purification of phage were carried out according to Studier (1969), and included a high salt-polyethylene glycol concentration step (Yamamoto et a/., 1970). Gene isolation and cloning T7 DNA was obtained by gentle phenol extraction of viral particles, followed by an additional purification (to remove RNA) by banding on CsCl density gradients. The DNA was n-butanol extracted to remove ethidium,

AND NICHOLSON

then dialyzed, phenol-extracted, and ethanol-precipitated. The 0.7 gene and specific subfragments therefrom were amplified by PCR using a Coy (Ann Arbor, MI) temperature cycler unit. DNA primers (10 pmol each) were combined with T7 DNA (4 fmol) in 100~~1 volumes containing buffer and 2.5 units of Vent DNA polymerase, heated at 95” for 1 min, then incubated at 50” for 1 min. DNA synthesis was then allowed for 5 min at 72”, and the cycle was repeated 31 times. The reaction products were electrophoresed on 1.2% agarose gels, and the DNA was purified from excised gel bands by phenol extraction and ethanol precipitation. The amplified 0.7 gene, or gene fragment was treated with Ndel, phenol-extracted, then digested with BarnHI. The BarnHI was removed by phenol extraction and the DNA gel-purified. The DNAs were ligated to gel-purified, No’ellBamHI-cut (see above) PET-11a DNA, and aliquots were used to transform competent HMS174 cells. Ampicillin-resistant colonies were isolated, and recombinant plasmids were identified which were introduced into HMSl74(DE3) cells. The recombinant plasmid bearing the entire T7(JS78) 0.7 gene was designated PET-JS78. Plasmids PET-PK(JS78) and PET-PK(A23) contained the protein kinase (PK) domain (see below) of the T7(JS78) and T7(A23) 0.7 genes, respectively. ln vivo radiolabeling of phage-infected and plasmid-containing cells The radioisotopic labeling of phage-infected, or IPTG-treated cells was performed essentially as described previously (Robertson and Nicholson, 1990). Cells were grown at 30” in low-phosphate minimal medium plus 0.4% glucose (Studier, 197313)to a density of 5 X 1O* cells/ml, then phage was added at an m.o.i. of 10. Alternatively, IPTG was added (final concentration 0.4 mn/r)to plasmid-containing cells grown to the same density. For double 35S/32P-labeling of cells, [35S]Express mix and inorganic [32P]phosphate (final concentration of each, 40 &i/ml) were added 2 min p.i., or 2 min following IPTG addition. Labeling reactions were stopped with an azide-phosphate quench, and RNasetreated extracts were prepared as described (Robertson and Nicholson, 1990). Gel electrophoretic

analysis of proteins

Nonradioactive proteins were electrophoresed in 12% polyacrylamide gels, containing 0.2% SDS as described (Studier, 1973b). Following electrophoresis, gels were stained with Coomassie brilliant blue, then destained, dried, and photographs were taken. Radioactive proteins were fractionated on lo-20% linear gradient polyacrylamide gels essentially as described

BACTERIOPHAGE

(Studier, 1973b). Following electrophoresis, the gels were dried directly and exposed to film. To differentiate the 35Sand 32P signals, the autoradiographic screening technique described by Cooper and Burgess (1982) was used. This provided a 32P autoradiogram completely free of 35S radioactivity, and an 35S autoradiogram essentially free of 32P radioactivity. RESULTS Expression in vivo of gp0.7-related from recombinant plasmids

polypeptides

The biochemical properties of gp0.7 suggested from the outset that the 0.7 gene would be difficult to clone in vectors that did not exert tight control over expression. In the absence of any evidence to the contrary, the gp0.7 protein kinase could be inimical to cell function, perhaps even when expressed in only catalytic amounts. However, gp0.7-dependent host shut off, which is rapidly imposed following 0.7 gene transcription (Rothman-Denes et al., 1973; McAllister and Barrett, 1977) would most certainly be toxic. A previous study noted the inability to clone DNA fragments containing the entire 0.7 gene (Studier and Rosenberg, 1981); however, there appears to be no &-acting inhibitory DNA sequences, since the same study showed that restriction fragments which spanned portions of the 0.7 gene could be stably cloned. We also were unsuccessful in obtaining stable recombinant plasmids containing the entire wild-type 0.7 gene, using a variety of different expression vectors (J. Michalewicz and A. W. Nicholson, unpublished experiments). It was clear that existing cloning vehicles were unable to reduce the basal expression of a toxic protein to levels that would permit isolation of stable recombinant plasmids containing the 0.7 gene. The route which proved successful entailed the PCR-mediated amplification of a specificT7 mutant 0.7 gene and its insertion into a highly repressible, T7 RNA polymerase-based expression vector. The plasmid PET-1 la was used, as it contains the lac operator directly downstream of the T7 410 promoter, and the /ac I gene, such that tight control over insert expression is obtained by formation of the lac repressor-operator complex, in repressor excess (Studier eta/., 1990). The application of PCR permitted the isolation of the 0.7 gene without flanking sequences, whose presence may have compromised the ability of the plasmid to control 0.7 expression [for example, a host promoter is approx. 12 bp downstream of the 0.7 gene stop codon in the T7 genome (Dunn and Studier, 1983)]. Moreover, to minimize nucleotide misincorporation during DNA amplification, Vent (or Tli) DNA polymerase (from T. littoralis) was used, which has a 3’-5’ exonuclease

T7 PROTEIN KINASE

455

(proofreading) activity (Ling et al., 1991; Cariello et al., 1991). To circumvent expression of the shut-off function, the 0.7 gene was isolated from T7(1S78) phage, which expresses a gp0.7-related polypeptide possessing onlythe protein kinase activity(Studier, 1973a; Robertson and Nicholson, 1990). Three types of PET-0.7 recombinants were prepared: plasmid PET-JS78 contained the entire T7(JS78) 0.7 gene; plasmid PETPK(JS78) contained the first 730 bp of the T7(JS78) 0.7 gene, encoding the protein kinase (PK) domain (see below), and plasmid PET-PK(A23) contained the corresponding DNA fragment from the T7(A23) 0.7 gene. The latter plasmid was prepared to examine the biochemical properties of the protein kinase domain of gp0.7(A23), which is incapable of phosphorylating its substrates in T7(A23)-infected cells (Studier, 1973a; Robertson and Nicholson, 1990).4 The identities of the cloned inserts were verified by size and restriction analysis. In addition, plasmid PET-PK(A23) was subjected to DNA sequence analysis, and of the approximately 400 bp sequenced, was found to contain the expected G to A transition as the only deviation from the wildtype sequence. This change creates the unique EcoRl site and abolishes a Hinfl site (see Fig. 5), which was verified by EcoRl and Hinfl restriction digestion (data not shown). To assay for expression of gp0.7-related polypeptides, HMS174(DE3) cells containing the PET-0.7 recombinant plasmids were grown in low phosphate minimal media containing glucose and antibiotic, and plasmid insert expression was induced during midexponential growth by the addition of IPTG. Upon continued incubation, the bacterial growth rates declined in all cases (data not shown). The retarded growth rate of IPTG-treated cells containing PET-1 1a, or the PET recombinants, agrees with previous observations (Studier et al., 1990). However, we did not observe any difference in the growth rates of IPTG-treated cells containing PET-1 1a, PET-PK(JS78) or PET-PK(A23), suggesting that expression of the protein kinase activity (see below) is not noticeably inhibitory to cell growth under these conditions. Proteins from cells containing the PET-0.7 recombinants were analyzed by gel electrophoresis (Fig. 2). The Coomassie-stained protein patterns revealed an approximately 30-kDa polypeptide (indicated by the arrow in Fig. 2) which was produced in IPTG-treated

4 Identical 1 D and 2D PAGE patterns of ‘“P-labeled phosphoproteins from T7(A23)-infected and T7(Cl 13) (which has a deleted 0.7 gene)-infected cells (Robertson and Nicholson, 1990; E. S. Robenson and A. W. Nicholson, unpublished experiments) provide the best evidence for the lack of gp0.7(A23) protein kinase actrvity in infected cells.

456 1

MICHALEWICZ 2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

M

AND NICHOLSON

reduced mobility is consistent with the nature of the A23 mutation, which results in a loss of negative charge in the PK domain of gp0.7 (see below). Plasmid-expressed gp0.7-dependent phosphotylation in viva

FIG. 2. Production of gp0.7.related polypeptides in recombinant pET plasmid-containing cells. HMSl74(DE3) ceils containing PET1 la, PET-JS78, PET-PK(JS78), or PET-PK(A23) were grown in LB broth at 30”, and at a density of 5 X 1O8 cells/ml were treated with 0.4 mM IPTG. Portions were taken 1, 2, and 3 hr following induction, and the proteins were analyzed on a 12% polyacrylamide gel containing 0.2% SDS. Proteins were visualized with Coomassie stain. Lanes l-4 show the time course for protein production in PET-1 1acontaining cells. Lane 1, no IPTG; lane 2, 1 hr; lane 3, 2 hr; lane 4, 3 hr following IPTG addition. Lanes 5-8 display the corresponding analysis of proteins from cells containing PET-JS78. Lanes 9-l 2 display the corresponding analysis of proteins from cells containing PET-PK(A23). Lanes 13-l 6 show proteins from IPTG-treated cells containing PET-PK(JS78). Lane 17 displays prestained protein size markers. The arrow indicates the position of the 30-kDa gpO.7-related polypeptide (see Results).

cells containing PET-JS78, PET-PK(JS78) or PETPK(A23), but not in IPTG-treated cells containing PET1 la. Densitometric analysis of individual gel lanes in Fig. 2 indicated that the gp0.7-related polypeptide represented approximately 5% of total cell protein, 3 hr following induction. The 30-kDa polypeptide appears to be the same as the approximately 28-kDa gpO.7-related species produced in T7(JS78)-infected cells (Studier and Rosenberg, 1981). The C-terminus of each of the polypeptides is established by the JS78 mutation, which creates an amber codon at position 243 in the gp0.7 amino acid sequence (see below). We conclude from these results that PET-JS78, PET-PK(JS78),and PET-PK(A23)can express the 26.8-kDa (based on the predicted amino acid sequence) N-terminal-containing fragment of gp0.7.5 The gp0.7-related polypeptide expressed from PET-PK(A23)exhibits a slightly lessened electrophoretic mobility, compared to the polypeptide expressed from PET-JS78 or PET-PK(JS78) (Fig. 2, compare lanes 1O-l 2 with lanes 6-8 and 14-l 6). The 5 It is currently unclear why the gp0.7.related polypeptides exhibit a larger Mr (30 kDa) than that predicted from the amino acid sequence (26.8 kDa); the anomolous electrophoretic mobility in SDScontaining polyacrylamide gels may reflect a structured polypeptide.

protein

The expression in viva of gp0.7-related polypeptides from the recombinant plasmids prompted the question whether the protein kinase activity is also expressed and, if so, whether the phosphoproteins produced are the same as those resulting from T7 infection. We anticipated that the protein phosphotransferase would be most active shortly following IPTG addition, since previous studies revealed that the protein kinase is most prominently expressed 8-l 1 min following infection and is subsequently inhibited, perhaps by gp0.7 phosphorylation (Pai et al., 1975b; Robertson and Nicholson, 1990). Thus, the gp0.7 protein kinase activity was looked for in PET-0.7 recombinant plasmid-containing cells several minutes as well as several hours following IPTG addition. To correlate the 32P-labeled phosphoprotein pattern with the total protein pattern, infected cells were incubated with inorganic [32P]phosphate and [35S]Express labeling mix, and differential autoradiography (Cooper and Burgess, 1982) applied. The results of this experiment are shown in Fig. 3. The 35S-labeled protein patterns are shown in (A) and the corresponding 32P-labeled protein patterns in (B). Similar to the unlabeled protein analysis, the gp0.7-related polypeptide is prominently featured as a 30-kDa 35S-labeled species 3 hr following IPTG treatment of cells containing PET-JS78, PET-PK(JS78), or PET-PK(A23) (Fig. 3A; lanes 9, 12, and 15; indicated by the arrow); smaller amounts of this protein are also detectable at shorter expression times (Fig. 3A, lanes 8, 11, and 14). As expected, this protein is not present in IPTG-treated cells containing PET-1la (Fig. 3A, lane 6). The 35S-labeled protein patterns of T7(JS78)-infected cells (Fig. 3A, lanes l-3 [HMS174(DE3)] and lanes 16-18 r$V31lo]) do not exhibit the gp0.7-related polypeptide, due to inherently lower levels of expression. Finally, there also occurs an approximately 25-kDa species, which is produced in all plasmid-bearing cells treated with IPTG which is apparently unrelated to gp0.7 (Fig. 3A, lanes 6, 9, 12, and 15). The gp0.7 phosphotransferase activity is expressed in IPTG-treated cells containing PET-JS78 or PETPK(JS78)(Figs. 3B and 3C). In particular, the 165-kDa 8’ subunit of RNA polymerase (Species l), and the 115 kDa polypeptide (Species 2) (Fig. 3B, lanes 8 and 15; and Fig. 3C, lanes 2-4, indicated by the arrows) are phosphorylated following gpO.7(JS78) production. These proteins were previously shown to be reliable

BACTERIOPHAGE

T7 PROTEIN KINASE

457

-2

FIG. 3. Production of phosphoproteins in cells expressing gp0.7-related polypeptides. HMS174(DE3) cells, with or without plasmid, were grown in minimal media plus 0.4% glucose (Studier, 1973b) at 30” to a density of 5 x 10’ cells/ml, and IPTG was added (final concentration 0.4 mM). Alternatively, cells were infected with T7(JS78) orT7(A23) phage at an m.o.i. of 7.5. Either 2 min or 3 hrfollowing IPTG induction, cells were labeled with inorganic [32P]phosphate (40 &i/ml) and [35S]Express (40 &i/ml) for 15 min; infected cells were labeled similarly from 2-l 7 min p.i. Cell extracts were prepared for analysis as described under Materials and Methods, and the proteins were analyzed by electrophoresis in a lo-20% gradient polyactylamide gel, which was directly dried and subjected to 35S and 32P analysis. (A) 35S-labeled protein patterns. (B) corresponding 3ZP-labeled protein patterns. Lanes l-3, HMSl74(DE3) cells, either uninfected (lane l), T7(A23)-infected (lane 2) or T7(JS78)-infected (lane 3). Lanes 4-6, PET-1 la-containing cell extracts; lane 4, uninduced; lane 5, 2 min postinduction; lane 6, 3 hr postinduction. Lanes 7-9, the corresponding analysis of PET-PK(A23)containing cell extracts; Lanes 13- 15, the corresponding analysis of PET-PK(JS78)containing cell extracts. Lane 16, T7(JS78)-infected; lane 17, T7(A23)-infected; lane 18, uninfected W3110 cell protein patterns. (C) In this experiment, HMS174(DE3) cells were labeled with 32P only, as described above. Proteins were analyzed from lane 1, uninfected cells containing PET-1 la; lane 2, T7(WT)-infected cells containing PET-1 la; lane 3, IPTG-treated cells containing PET-JS78; lane 4, IPTG-treated cells containing PETPK(JS78); lane 5, IPTG-treated cells containing PET-PK(A23); and lane 6, T7(A23)-infected cells containing PET-1 la. In all panels, “1” and “2” refer to protein species 1 and 2 (see text); the lower, larger arrow indicates the position of the 30-kDa gp0.7.related polypeptide.

indicators of gp0.7 protein kinase activity, using the 1 D SDS-PAGE assay (Robertson and Nicholson, 1990). The two 32P-labeled proteins overlap with 35S-labeled protein bands (Fig. 3A, indicated by the arrows), which in particular shows that the p’ subunit carries the 32P-radioactivity. The PET-JS78 and gp0.7-dependent PET-PK(JS78)-dependent phosphoprotein patterns are

essentially identical to that produced by T7(JS78) infection (Fig. 3B, compare lanes 3 and 16 with lanes 8 and 14). The phosphoprotein pattern from IPTG-treated cells containing PET-1 1a (Fig. 3B, lane 5) is essentially identical to that of uninfected cells (Fig. 3B, compare lanes 1 and 18 with lane 4, and lanes 2 and 17 with lane 11). Under these experimental conditions, the

458

MICHALEWICZ

gp0.7 protein kinase activity is more strongly expressed in phage-infected cells than in IPTG-treated, recombinant plasmid-bearing cells; however, we have not yet established conditions for optimal expression of the protein kinase activity from the plasmids. The plasmid-encoded gp0.7-related polypeptides undergo phosphorylation (Fig. 3B, lanes 9, 12, and 15; Fig. 3C, lanes 3 and 4; indicated by the lower arrow). Little, if any, transphosphorylation is observed at longer expression times (Fig. 3B, lanes 9, 12, and 15) despite the substantial levels of gp0.7-related polypeptide (Fig. 3B, lanes 9, 12, and 15). The production of cell proteins is also diminished in this situation, so it cannot be determined whether transphosphorylation is inhibited by gp0.7 phosphorylation, as suggested earlier (Pai et a/., 197513). Interestingly, the gp0.7-related polypeptide expressed from PET-PK(A23) is also phosphorylated (Fig. 3B, lane 9; Fig. 3C, lane 5). Assuming that gp0.7 phosphorylation is self-catalyzed, it follows that the A23 mutation affects neither ATP-binding nor catalysis of phosphate transfer per se. Moreover, it is also apparent that expression of the gp0.7-related mutant polypeptide from PET-PK(A23) results in phosphorylation of the p’ subunit and, to a lesser extent, species 2 (Fig. 3C, lane 5). This is in contrast to gpO.7(A23) expressed during phage infection, which fails to cause phosphorylation (Fig. 3C, lane 6) and, in contrast to the experiment using PET-PK(JS78), which shows preferential phosphorylation of species 2. Thus, the A23 mutation does not inhibit, but rather alters the pattern of phosphorylation produced by the gp0.7-related polypeptide, as expressed from a plasmid (see also below and Discussion). Plasmid-expressed phosphotransferase

gp0.7 protein activity in vitro

Previous studies demonstrated that gp0.7 synthesized in vitro or in vivo can catalyze the phosphorylation of exogenous proteins in vitro (Rahmsdorf et a/., 1974; Pai et al., 197513). We therefore examined gp0.7 protein kinase activity in extracts of IPTG-treated, PET0.7 recombinant plasmid-containing cells. The assay utilized egg white lysozyme, previously shown to be a good substrate for the gp0.7 protein kinase (Pai et a/,, 1975b). Extracts of T7(JS78)- or T7(A23)-infected cells were also tested. The results are presented in Table 1, where it is seen that gp0.7 protein kinase activity is present in extracts of IPTG-treated cells carrying PETJS78 or PET-PK(JS78). The levels of phosphotransferase activity are similar in both cases. However, the level of expression of the gp0.7 protein kinase in PETJS78-containing cell extracts is about 50% the level obtained with T7(JS78)-infected cell extracts. There are

AND NICHOLSON TABLE 1 17 ~~0.7 PROTEINKINASEACTIVITYIN EXTRACTSOF CELLS CONTAINING PET-0.7 RECOMBINANTPLASMIDSOR INFECTEDWITH PHAGE

Cell extract sourcea

0.4 mM IPTG

Lysozyme substrate*

+ + + + + + + + + +

1 2

HMSl74(DE3) HMSl74(DE3)

-

3 4

HMSl74(DE3)-T7(JS78) HMSl74(DE3)-T7(JS78)

-

5 6

HMSl74(DE3)-T7(A23) HMS174(DE3)-T7(A23)

-

7 8

HMSl74(DE3)/pET-1 HMSl74(DE3)/pET-1

1A 1A

+ +

9 10 11

HMSl74(DE3)/pET-JS78 HMS 174(DE3)/pET-JS78 HMSl74(DE3)/pET-JS78

+ + -

12 13 14

HMS174(DE3)/pET-PK(JS78) HMS174(DE3)/pET-PK(JS78) HMS174(DE3)/pET-PK(JS78)

+ + -

15 16 17

HMSl74(DE3)/pET-PK(A23) HMSl74(DE3)/pET-PK(A23) HMSl74(DE3)/pET-PK(A23)

+ + -

fmol 32P transferred 9.8 29.0 8.8 229.0 10.0 44.5 26.7 41.0 1 1.4 105.0 32.7 8.6 132.0 26.5 12.9 362.0 29.2

a To prepare cell-free extracts, HMSl74(DE3) cells were grown in minimal media plus 0.4% glucose (Studier, 1973b) at 30” to a density of 5 X 10’ cells/ml (50 ml total volume). Purified phage were then added at an m.o.i. of 7.5 and cells collected 10 min p.i. Alternatively, plasmid-containing HMS174(DE3) cells were grown to the same density, then treated with IPTG (0.4 mM final concentration) and cells collected 20 min thereafter. Extracts were prepared according to Godson and Sinsheimer (1967). * Protein kinase assays were performed according to Pai et a/. (1975a) using egg white lysozyme as substrate. Reaction mixtures (100 ~1) contained 30 ~1 of cell extract, 10 &i of [Y-~‘P]ATP, and substrate (1.3 fig/PI) where indicated. Following incubation at 30” for 30 min, reactions were stopped by addition of 10% trichloroacetic acid, heated at 100’ for 15 min and filtered through Whatman 3M filters. The filters were batch-washed three times, dried, then counted in scintillation fluid. Assays were performed in duplicate, which agreed to within 1 1%; 10 fmol corresponds to 510 dpm. In separate experiments, using extracts containing active gp0.7 protein kinase, the rncorporation of 32P radioactivity into lysozyme was linear up to 40 min at 30”. One-dimensional SDS-PAGE analysis of the kinase reactions showed that 32P radioactivity was incorporated into the 14-kDa lysozyme (data not shown). Also, the Coomassiestained gel patterns revealed the production of gp0.7.related polypeptide in extracts of IPTG-induced, recombinant plasmid-contarning cells. The background (minus lysozyme) levels of kinase activity were probably slightly higher than normal, since the cell extract preparation included a lysozyme treatment step.

only negligible amounts of phosphotransferase activity in extracts of uninfected cells, or of cells containing PET-1 1a (see also note to Table 1). Interestingly, there is substantial phosphorylation of lysozyme in extracts of PET-A23(PK)-containing cells, whereas there is only

459

BACTERIOPHAGE T7 PROTEIN KINASE

a minor amount of phosphorylation in the T7(A23)-infected cell extracts. This result is consistent with the assertion (see above) that the A23 mutation, when present in the truncated form of gp0.7 expressed from a plasmid, alters the interaction of the protein kinase with its substrates (see Discussion).

ing PET-JS78grow slightly more slowly than cells containing PET-PK(JS78), perhaps suggesting a minor amount of nonsense suppression in HMS174(DE3) cells (J. Michalewicz and A. W. Nicholson, unpublished experiments). DISCUSSION

Sequence analysis of the 0.7 gene A23 and JS78 mutations Identification of the 0.7 gene A23 and JS78 mutations could provide insight into the mechanisms of gp0.7-catalyzed phosphate transfer and host shut off. Nucleotide sequence analysis of the T7(A23) and T7(JS78) 0.7 genes was performed using purified viral DNA as the sequencing template. Plasmid PETPK(A23) was also similarly analyzed. The choice of sequencing primers was guided by the approximate locations of the mutations, determined in previous genetic mapping studies (Studier, 1973a; Simon and Studier, 1973; Studier et al., 1979; Studier and Rosenberg, 1981). Specifically, the A23 mutation occurs within the first 28% of the 0.7 gene, while the JS78 mutation maps closer to the 3’ end of the gene, at approximately 679/o. The results of T7(A23) and PET-PK(A23) sequence analysis revealed a single GC to AT bp transition at nucleotide position 2318, which was the only difference between mutant and wild-type sequences in this region. The predicted amino acid change is an Asp to Asn substitution at position 100 (Fig. 4A), which is directly downstream of the putative ATP-binding domain (underlined in Fig. 4A), and is close to a highly conserved lysine residue, implicated in the catalysis of phosphate transfer (Hanks et al., 1988). The formal loss of a negative charge in gp0.7 incurred by the A23 mutation may explain the slower gel electrophoretic mobility of the gp0.7(A23)-related polypeptide, compared to gp0.7(JS78)-related polypeptide (see above). Sequence analysis of the JS78 mutation revealed a CG to TA transition at nucleotide position 2747, creating an amber codon in place of Gln 243 (Fig. 48). This provides an explanation for the production of a 30-kDa gp0.7 polypeptide, rather than the 41-kDa wild-type protein, in T7(JS78)-infected, or PET-JS78-containing cells. The existing biochemical and genetic evidence (see Discussion) implicates the C-terminal 13-kDa domain in host shut off; the absence of this segment in gpO.7(JS78) explains the inability of T7(JS78) to exert shut off (E. S. Robertson and A. W. Nicholson, unpublished experiments). The viability of cells containing PET-JS78 indicates that the amber nonsense codon is efficiently recognized, since any appreciable readthrough would result in expression of the toxic shut-off activity. However, we have observed that cells contain-

This report has described the plasmid cloning of the phage T7 0.7 (protein kinase) gene and expression of the gp0.7 protein kinase activity. The cloning strategy was based in part on the assumption that gp0.7-dependent shut off is a toxic function, and that plasmids containing the wild-type 0.7 gene, or 0.7 subfragments which encode this activity would be unable to be stably maintained. Our efforts were therefore directed toward obtaining a clone of the T7(JS78) 0.7 gene, which only expresses the protein kinase activity. We had originally inserted the T7(JS78) 0.7 gene into plasmid pDOC55, a pUC18 derivative designed to exert tight control over insert expression through utilization of an antisense transcript, as well as the cl repressor-controlled P, promoter (O’Connor and Timmis, 1987). However, the gp0.7-related polypeptide was not produced in cells containing a 0.7(JS78)-pDOC55 recombinant plasmid, using a variety of means to induce expression (J. Michalewicz and A. W. Nicholson, unpublished experiments). The PET plasmid system (Studier et al., 1990) readily permitted the cloning, in expressible form, of the T7(JS78) 0.7 gene, as well as a specific subfragment encoding the protein kinase domain. Production of the plasmid-encoded, gp0.7-related polypeptide was IPTG-dependent, and cell growth following induction was not inhibited over and above that of nonrecombinant plasmid-containing cells. Expression in viva of the gp0.7-related polypeptide from plasmid PETJS78 or PET-PK(JS78)results in the phosphorylation of the same proteins as those modified following T7 infection. Thus, the gp0.7 protein kinase does not require an additional viral function(s) for its specific action. The gp0.7-related polypeptides expressed from the recombinant plasmids undergo phosphorylation in vivo; however, to determine whether gp0.7 phosphorylation is self-catalyzed will require in vitro assays using purified gp0.7. Perhaps a mutant gp0.7 can be obtained which is resistant to self-phosphorylation, allowing the expression and purification of a protein kinase which is not subject to possible self-regulation. The gp0.7-related polypeptide encoded by PETPK(A23) exhibits an altered protein kinase activity. The mutant protein kinase fragment preferentially phosphorylates the RNA polymerase p’ subunit in viva and is able to phosphorylate lysozyme in w’rro. This is in marked contrast to viral-encoded gp0.7(A23), which

460

MICHALEWICZ

A

AND NICHOLSON

A23 Mutation

f240

A

T280

G 74

N

G

H

F

S

A

A

Y

S

H

P

L

L

P

N

R

60

V

I

K

90

V

G

F

K

l

K

E

D,ooS

G

1

I

I

2320 ’

t

GGT-AAT-GGT.CAC-TTC-TCG-GCT-GCT-TAT-AGT-CAC-CCG-CTG-CTA-CCT-AAC-AGA-GTG-ATT-MG-GTC-GGC-m-MG-~-GAG-GAT-TCA-GGC-

N

B

JS76

Mulatlon

f750 3790 I CGA-CAG-AAA-GAA-A~-GAC-CGC-GCT-A*G-GCC-CGT-AAA-GAA-CGT-CAC-GAG-GGG-CGC-TTA-GAG-GCA-CGC-AGA-~C-AAA-CGT-CGC RQKElDRAKAAKERHEGRLEARRFKRR 242 4 I STOP

250

260

2630 I MC-CGC-MG-GCA-CGT-AAA-GCA-CAC-AAA-GCT-AAG-CGC-GAA-AGA-ATG-~= NRKARKAHKAKRERML270

I

2670 -

260

FIG. 4. Identification of the 0.7 gene A23 and JS78 mutations. (A)The nucleotide sequence and predicted amino acid sequence of the region of gp0.7 near the A23 mutation, which is a G to A change at nucleotide position 23 18, causing an Asp to Asn change at amino acid 100. The underlined region refers to the putative ATP-binding domain, which includes the conserved lysine (Hanks et a/., 1988) as indicated by the asterisk. (5) The nucleotide sequence and predicted amino acid sequence of the region of gp0.7 near the JS78 mutation, which is a C to U change at nucleotide position 2747, forming an amber nonsense codon. The underlined region indicates the strongly basic amino acid segment, implicated in shut off (see Discussion).

fails to phosphorylate substrates in viva or in vitro. The difference in biological activities may result from the fact that the plasmid-encoded gp0.7(A23) lacks the Cterminal 1 l-kDa domain, whose presence may inhibit protein kinase activity of the mutant protein. The biochemical behavior of plasmid-encoded gp0.7(A23) suggests that Asp1 00 may play a role in substrate recognition; since the mutation (Asp to Asn) results in a loss of negative charge, a prediction would be that the recognition motif in the gp0.7 protein kinase substrates contains a basic residue(s). Further biochemical analysis of gp0.7(A23), as well as other gp0.7 variants, and the characterization of substrate phosphorylation sites should allow a definition of the specificity and catalytic determinants for the gp0.7 phosphotransferase activity. We were unable to clone either the entire wild-type 0.7 gene, or a subfragment encoding the C-terminal domain, important for shut off (see below). These negative results, although not conclusive, support the assertion that gp0.7 shut-off activity is a strongly toxic function. It may be possible to express shut-off activity from PET-JS78 by the controlled production of an Su2 suppressor, allowing read-through synthesis of the 4 lkDa wild-type protein [the natural termination signal for the 0.7 cistron is UGA (Dunn and Studier, 1983)]. The ability to overexpress wild-type gp0.7, however, may

only be realized if the probable concomitant feedback repression of host transcription can be circumvented. The global inhibitory effect of gp0.7 on host transcription prompts the question of the shut-off mechanism. The approximately 1 1-kDa C-terminal domain of gp0.7 is necessary and sufficient for shut off, since gpO.7(JS78) fails to exert shut off and since several mutant gp0.7 fusion proteins which bear the N-termini of upstream gene products are active in shut off (Rothman-Denes et al., 1973; Simon and Studier, 1973; McAllister and Barrett, 1977; Studier et al., 1979). Also, studies of the uv-induced transcriptional inactivation of T7 early gene expression (Herrlich et a/., 1974) indicated that, compared to the protein kinase domain, the shut-off domain is distal to the T7 early host promoters. Examination of the predicted amino acid sequence of the C-terminal portion of gp0.7 reveals a strongly basic region (63% Lys, Arg, His), encompassing about 30 amino acids (see underlined sequence in Fig. 5B). A protein sequence databank search revealed a close similarity of the basic segment with protamines -highly basic polypeptides that tightly bind DNA and are involved in condensing eukaryotic chromatin into a transcriptionally inactive state (Bloch, 1969). It is therefore possible that this portion of gp0.7 binds DNA, providing the first step in catalysis of shut-off. There is some indirect experimental evidence in support of this

BACTERIOPHAGE

model (Hodgson et al., 1985; Pai et a/., 1975; Zillig et a/., 1975), including the observation that the E. co/i B nucleoid undergoes deformation, then condenses shortly after T7 infection (Luria and Human, 1950). A previous report described the presence of a protamine-like protein in E. co/i (Altman et al., 1981). This 33 amino acid polypeptide is of comparable size to the basic segment of gp0.7, but there is no clear sequence similarity. The results of this study support a model of gp0.7 which has two independent functional domains, each expressing an activity involved in different aspects of T7 reproduction. Perhaps the 0.7 gene was created from the fusion of a shut-off gene with a protein kinase gene, as a consequence of selective pressure to maximize the coding capacity of a size-limited genome. A protein sequence database search indicated that the 27-kDa protein kinase domain does not exhibit any obvious similarity with other prokaryotic protein kinases, with the possible exception of an aminoglycoside phosphotransferase. Interestingly, other than the conserved ATP-binding region (see Fig. 4), the gp0.7 protein kinase does not exhibit any of the conserved sequence elements seen in eukaryotic serine/threoninespecific protein kinases (Hanks et a/., 1988). One can speculate that the gp0.7 protein kinase was originally a cell-encoded activity which was incorporated into a T7 progenitor. A prediction would be that a protein kinase with a similar functional role(s) exists in a cellular system. Toward these ends, an assessment of the influence of gp0.7-catalyzed protein phosphorylation on cell function may be permitted through the controlled expression of the protein kinase from a low copy number plasmid, or from the host chromosome. These avenues of experimentation and others should provide a detailed picture of the role of a protein kinase in the strategy of viral infection.

ACKNOWLEDGMENTS We thank Bill Studier for provision of T7 mutants, Alan Rosenberg (Brookhaven National Laboratory) for suggesting the use of the PET vector/PCR cloning strategy, and Scott Goustin (Center for Molecular Biology, Wayne State University) for initial advice on application of PCR techniques. We also thank Lee Aggison for carrying out DNA sequence analyses. The assistance and ongoing interest of Erle S. Robertson In these studies are appreciated, as well as the comments of an anonymous reviewer regarding the analysis of the A23 mutation. The Macromolecular Analysis Core Facility of Wayne State University is acknowledged for synthesis of the DNA ollgonucleotides. This work was supported in part by a grant from the National Institutes of Health (GM41 283), by Grant IN-1 62 from the American Cancer Society, and also by the Center for Molecular Biology, Wayne State University.

REFERENCES ALTMAN, S., MODEL, P., DIXON, G. H., and WOSNICK, M. A. (198 1). An E. co/i gene coding for a protamine-like protein. Cell 26, 299-304.

T7 PROTEIN KINASE

461

ATKINSON, T., and SMITH, M. (1984). “Oligonucleotide Synthesis: A Practical Approach” (M. J. Gait, Ed.), pp. 35-81. IRL Press, Oxford. BLOCH, D. P. (1969). A catalog of sperm histones. Generics Sup@. 61, 93-111. BRUNOVSKIS,I., and SUMMERS, W. C. (1971). The process of infection with coliphage T7. V. Shutoff of host RNA synthesis by an early phage function. Virology 45, 224-231. BRUNOVSKIS,I., and SUMMERS,W. C. (1972). The process of infection with coliphage T7. VI. A phage gene controlling shutoff of host RNA synthesis. Virology 50, 322-327. CAMPBELL, J. L., RICHARDSON,C. C.. and STUDIER, F. W. (1978). Genetic recombination and complementation between bacteriophage T7 and cloned fragments of T7 DNA. Proc. Nat. Acad. SC;. USA 75, 2276-2280. CARIELLO, N. F., SWENBERG,J. A., and SKOPEK,T. R. (1991). Fidelity of Thermococcus littoralis DNA polymerase (Vent) in PCR determined by denaturing gradient gel electrophoresis. Nucleic Acids Res. 19,4193-4198. CHAMBERLIN, M., and RYAN, T. (1982). Bacteriophage DNA-dependent RNA polymerases. In “The Enzymes” (P. D. Boyer, Ed.), pp. 87-l 08. Academic Press, New York. COOPER, P. C., and BURGESS,A. W. (1982). Simultaneous detection of %- and 32P-labeled proteins on electrophoretic gels. Anal. Biochem.126,301-305. CORTAY, J-C., RIEUL, C., DUCLOS, B. and COZZONE. A. J. (1986). Characterization of the phosphoproteins of Escherichia co/i cells by electrophoretic analysis. Eur. J. Biochem. 159, 227-237. DUNN, J. J.. and STUDIER, F. W. (1983). Complete nucleotide sequence of bacteriophage T7 DNA and the locations of T7 genetic elements. J. Mol. Biol. 166, 477-535. ENAMI, M., and ISHIHAMA, A. (1984). Protein phosphorylation in Escherichia co/i and purification of a protein kinase. 1. Biol. Chem. 259, 526-533. GODSON, G. N., and SINSHEIMER, R. L. (1967). Use of Brij lysis as a general method to prepare polyribosomes from Escherichia co/i. Biochim. Biophys. Acta 149, 489-495. GOMEZ, B., and NUALART. L. (1977). Requirement of the bacteriophage T7 0.7 gene for phage growth in the presence of the Col lb factor. J. Gen. Viral. 35, 99-106. HANKS, S. K.. QUINN, A. M., and HUNTER,T. (1988). The protein kinase family: Conserved features and deduced phylogeny of the catalytic domains. Science (USA) 241,42-52. HAUSMANN, R. (1988). The T7 group. In “The Bacteriophages” (R. Calendar, Ed.), pp. 259-289. Plenum, New York. HERRLICH,P., RAHMSDORF,H. J., PAI, S. H., and SCHWEIGER.M. (1974). Translational control induced by bacteriophage T7. froc. Nat. Acad. Sci. USA 71, 1088-l 092. HESSELBACH,B. A., and NAKADA, D. (1977). “Host shutoff” function of bacteriophage T7: Involvement of T7 gene 2 and gene 0.7 in the inactivation of fscherichia co/i RNA polymerase. 1. Viral. 24, 736745. HIRSCH-KAUFFMANN,M., HERRLICH,P., PONTA, H., and SCHWEIGER,M. (1975). Helper function of T7 protein kinase In virus propagation. Nature (London) 255, 508-510. HODGSON, D., SHAPIRO, L., and AMEMIYA, K. (1985). Phosphotylation of the 0 subunit of RNA polymerase and other host proteins upon &dl infection of Caulobacter crescentus. Virology 55, 238-241. HUBER, H. E., BEAUCHAMP,B. B., and RICHARDSON,C. C. (1988). fscherichia co/i dGTP triphosphohydrolase is inhibited by gene 1.2 protein of bacteriophage T7.1. Biol. Chem. 263, 13,549-l 3,556. ITIKAWA,H., WADA, M., SEKINE, K., and FUJITA, H. (1989). Phosphory lation of glutaminyl-tRNA synthetase and threonyl-tRNA synthe-

462

MICHALEWICZ

tase by the gene products of dnaK and dnal in Escherichia co/i K 12 cells. Biochimie 71, 1079-l 087. ITO, K., EGAWA, K., and NAKAMURA, Y. (1991). Genetic interaction between the @’ subunit of RNA polymerase and the arginine-rich domain of Escherichia coli NusA protein. J. Bacterial. 173, 14921501. KRUGER, D. H., and SCHROEDER,C. (1981). Bacteriophage T3 and bacteriophage T7 virus-host cell interactions. Microbial. Rev. 45, 9-51. LING, L. L., KEOHAVONG,P., DIAS, C., and THILLY, W. G. (1991). Optimization of polymerase chain reaction with regard to fidelity: Modified T7, Taq and Vent DNA polymerases. PCR Methods Appl. 1, 63-69. LIJRIA,S. E., and HUMAN, M. L. (1950). Chromatin staining of bacteria during bacteriophage infection. J. Bacterial. 59, 551-560. MASAMUNE, Y., FRENKEL, G. D., and RICHARDSON, C. C. (1971). A mutant of bacteriophage T7 deficient in polynucleotide ligase. J. Biol. Chem. 246, 6874-6879. MAYER, J. E., and SCHEWIGER,M. (1983). RNase III is positively regulated by T7 protein kinase. J. Biol. Chem. 258, 5340-5343. MCALLISTER, W. T., and BARRET, C. L. (1977). Roles of the early genes of bacteriophage T7 in shutoff of host macromolecular synthesis. J. Viral. 23, 543-553. MERTENS, H., and HAUSMANN. R. (1982). Coliphage BA14: A new relative of phage T7. J. Gen. Viral. 62, 331-341. O’CONNOR, C. D., and TIMMIS, K. N. (1987). Highly-repressible system for cloning genes that specify potentially toxic proteins. J. Batteriol. 169, 4457-4462. PAI, S-H., PONTA, H., RAHMSDORF,H-J., HIRSCH-KAUFMANN,M., HERRLICH, P., and SCHWEIGER,M. (1975a). Protein kinase of bacteriophage T7. 1. Purification. Eur. J. Biochem. 55, 299-304. PAI, S-H., RAHMSDORF,H-J., PONTA, H., HIRSCH-KAUFMANN, M., HERRLICH, P., and SCHWEIGER,M. (1975b). Protein kinase of bacteriophage T7. 2. Properties, enzyme synthesis in vitro and regulation of enzyme synthesis and activity in vivo. Eur. J. Biochem. 55, 305314. PFENNIG-YEH, M., PONTA, H., HIRSCH-KAUFFMANN, M., RAHMSDORF, H.-J., HERRLICH, P., and SCHWEIGER,M. (1978). Early T7 gene expression. Mol. Gen. Genet. 166, 127-140. PONTA. H., RAHMSDORF, H.-J., PAI, S.-H., HIRSCH-KAUFMANN, M., HERRLICH, P., and SCHWEIGER,M. (1974). Control of gene expression in bacteriophage T7: Transcriptional controls. MO/. Gen. Genet. 134,281-297. RAHMSDORF,H.-J., PAI, S.-H., PONTA, H., HERRLICH, P., ROSKOSKI,R., JR., SCHWEIGER,M., and STUDIER, F. W. (1974). Protein kinase induction in Escherichia coli by bacteriophage T7. Proc. Natl. Acad. Sci. USA 71,586-589.

AND NICHOLSON ROBERTSON.E. S., and NICHOLSON, A. W. (1990). Protein klnase of bacteriophage T7 Induces the phosphotylation of only a small number of proteins In the infected cell. Virology 175, 525-534. ROTHMAN-DENES, L. S., MUTHUKRISHNAN.S., HASELKORN,R.. and STUDIER, F. W. (1973). A T7 gene function required for shut-off of host and early T7 transcription. In “Virus Research” (C. F. Fox and W. S. Robinson, Eds.), pp. 227-239. Academic Press, New York. SIMON, M. N., and STUDIER, F. W. (1973). Physical mapping of the early region of bacteriophage T7 DNA. J. Mol. Biol. 79, 249-265. STUDIER,F. W. (1969). The genetics and physiology of bacteriophage T7. Virology 39, 562-574. STUDIER, F. W. (1973a). Genettc analysis of non-essential bacteriophage T7 genes. J. Mol. Biol. 79, 227-236. STUDIER,F. W. (1973b). Analysis of bacteriophage T7 early RNAs and proteins on slab gels. J. Mol. Biol. 79, 237-248. STUDIER, F. W. (1979). Relationships among different strains of T7 and among T7-related bacteriophages. Virology 95, 70-84. STUDIER, F. W. (1975). Gene 0.3 of bacteriophage T7 acts to overcome the DNA restriction system of the host. J. Mol. Biol. 94, 283-295. STUDIER, F. W., and DUNN, J. 1. (1983). Organization and expression of bacteriophage T7 DNA. Cold Spring Harbor Symp. Quant. Biol. 47, 999-l 007. STUDIER,F. W., and MOFFATT, B. A. (1986). Use of bacteriophage T7 to direct selective high-level expression of cloned genes. J. Mol. Biol. 189, 113-l 30. STUDIER, F. W., and MOWA, N. R. (1976). SAMase gene of bacteriophage T3 is responsible for overcoming host restriction. J. Viral. 19, 136-145. STUDIER, F. W., and ROSENBERG,A. H. (1981). Genetic and physical mapping of the late region of bacteriophage T7 DNA by use of cloned fragments of T7 DNA. J. Mol. Biol. 153, 503-525. STUDIER,F. W., ROSENBERG,A. H., DUNN, J. J., and DUBENDORFF,1. W. (1990). Use of T7 RNA polymerase to direct expression of cloned genes. ln “Methods in Enzymology” (D. V. Goeddel, Ed.), Vol. 185, pp. 60-89. Academic Press, New York. STUDIER, F. W.. ROSENBERG,A. H., SIMON, M. N., and DUNN, J. J. (1979). Genetic and physical mapping in the early region of bacteriophage T7 DNA. J. Mol. Biol. 135, 917-937. YAMAMOTO, K. R., ALBERTS. B. M., BENZINGER,R., LAWHORNE,L., and TREIBER,G. (1970). Rapid bacteriophage sedimentation in the presence of polyethylene glycol and its application to large-scale virus purification. Virology 40, 734-744. ZILLIG, W., FUJIKI.H., BLUM, W., JANEKOVIC,D., SCHWEIGER.M., RAHMSDORF, H.-J., PONTA, H., and HIRSCH-KAUFFMANN,M. (1975). ln vivo and in vitro phosphotylation of DNA-dependent RNA polymerase of Escherichia co/i by bacteriophage-T7induced protein kinase. Proc. Nat/. Acad. Sci. USA 72, 2506-25 10.